Study of parallel prediction of groundwater table by BP neural network

This paper proposes a suitable method of prediction of water table of water source field in confined aquifers in porous rocks. This prediction and optimization are done by BP-Genetic Geography cellular automata. The principle of this method is from the framework of geography cellular automata. Every field of water source is regarded as a cell. Scenario of next time of every field of water source is depended on the water source field itself and scenario of last time of the water source in the borderland and behavior of next time of water source in the borderland, as well as environmental conditions in the water source site.